Research Papers

As machine learning models are increasingly used in educational settings, from detecting at-risk students to predicting student performance, algorithmic bias and its potential impacts on students raise critical concerns about algorithmic fairness. Although group fairness is widely explored in education, works on individual fairness in a causal context are understudied, especially on counterfactual fairness. This paper explores the notion of counterfactual fairness for educational data by conducting counterfactual fairness analysis of machine learning models on benchmark educational datasets. We demonstrate that counterfactual fairness provides meaningful insight into the causality of sensitive attributes and causal-based individual fairness in education.

As artificial intelligence (AI) systems become increasingly embedded in critical societal functions, the need for robust red teaming methodologies continues to grow. In this forum piece, we examine emerging approaches to automating AI red teaming, with a particular focus on how the application of automated methods affects human-driven efforts. We discuss the role of labor in automated red teaming processes, the benefits and limitations of automation, and its broader implications for AI safety and labor practices. Drawing on existing frameworks and case studies, we argue for a balanced approach that combines human expertise with automated tools to strengthen AI risk assessment. Finally, we highlight key challenges in scaling automated red teaming, including considerations around worker proficiency, agency, and context-awareness.

Frontier AI models -- highly capable foundation models at the cutting edge of AI development -- may pose severe risks to public safety, human rights, economic stability, and societal value in the coming years. These risks could arise from deliberate adversarial misuse, system failures, unintended cascading effects, or simultaneous failures across multiple models. In response to such risks, at the AI Seoul Summit in May 2024, 16 global AI industry organizations signed the Frontier AI Safety Commitments, and 27 nations and the EU issued a declaration on their intent to define these thresholds. To fulfill these commitments, organizations must determine and disclose ``thresholds at which severe risks posed by a model or system, unless adequately mitigated, would be deemed intolerable.'' To assist in setting and operationalizing intolerable risk thresholds, we outline key principles and considerations; for example, to aim for ``good, not perfect'' thresholds in the face of limited data on rapidly advancing AI capabilities and consequently evolving risks. We also propose specific threshold recommendations, including some detailed case studies, for a subset of risks across eight risk categories: (1) Chemical, Biological, Radiological, and Nuclear (CBRN) Weapons, (2) Cyber Attacks, (3) Model Autonomy, (4) Persuasion and Manipulation, (5) Deception, (6) Toxicity, (7) Discrimination, and (8) Socioeconomic Disruption. Our goal is to serve as a starting point or supplementary resource for policymakers and industry leaders, encouraging proactive risk management that prioritizes preventing intolerable risks (ex ante) rather than merely mitigating them after they occur (ex post).

The development of artificial general intelligence (AGI) is likely to be one of humanity's most consequential technological advancements. Leading AI labs and scientists have called for the global prioritization of AI safety citing existential risks comparable to nuclear war. However, research on catastrophic risks and AI alignment is often met with skepticism, even by experts. Furthermore, online debate over the existential risk of AI has begun to turn tribal (e.g. name-calling such as "doomer" or "accelerationist"). Until now, no systematic study has explored the patterns of belief and the levels of familiarity with AI safety concepts among experts. I surveyed 111 AI experts on their familiarity with AI safety concepts, key objections to AI safety, and reactions to safety arguments. My findings reveal that AI experts cluster into two viewpoints -- an "AI as controllable tool" and an "AI as uncontrollable agent" perspective -- diverging in beliefs toward the importance of AI safety. While most experts (78%) agreed or strongly agreed that "technical AI researchers should be concerned about catastrophic risks", many were unfamiliar with specific AI safety concepts. For example, only 21% of surveyed experts had heard of "instrumental convergence," a fundamental concept in AI safety predicting that advanced AI systems will tend to pursue common sub-goals (such as self-preservation). The least concerned participants were the least familiar with concepts like this, suggesting that effective communication of AI safety should begin with establishing clear conceptual foundations in the field.

In recent years, much research has been dedicated to uncovering the environmental impact of Artificial Intelligence (AI), showing that training and deploying AI systems require large amounts of energy and resources, and the outcomes of AI may lead to decisions and actions that may negatively impact the environment. This new knowledge raises new ethical questions, such as: When is it (un)justifiable to develop an AI system, and how to make design choices, considering its environmental impact? However, so far, the environmental impact of AI has largely escaped ethical scrutiny, as AI ethics tends to focus strongly on themes such as transparency, privacy, safety, responsibility, and bias. Considering the environmental impact of AI from an ethical perspective expands the scope of AI ethics beyond an anthropocentric focus towards including more-than-human actors such as animals and ecosystems. This paper explores the ethical implications of the environmental impact of AI for designing AI systems by drawing on environmental justice literature, in which three categories of justice are distinguished, referring to three elements that can be unjust: the distribution of benefits and burdens (distributive justice), decision-making procedures (procedural justice), and institutionalized social norms (justice as recognition). Based on these tenets of justice, we outline criteria for developing environmentally just AI systems, given their ecological impact.

Machine-generated music (MGM) has become a groundbreaking innovation with wide-ranging applications, such as music therapy, personalised editing, and creative inspiration within the music industry. However, the unregulated proliferation of MGM presents considerable challenges to the entertainment, education, and arts sectors by potentially undermining the value of high-quality human compositions. Consequently, MGM detection (MGMD) is crucial for preserving the integrity of these fields. Despite its significance, MGMD domain lacks comprehensive systematic evaluation results necessary to drive meaningful progress. To address this gap, we conduct experiments on existing large-scale datasets using a range of foundational models for audio processing, establishing systematic evaluation results tailored to the MGMD task. Our selection includes traditional machine learning models, deep neural networks, Transformer-based architectures, and State space models (SSM). Recognising the inherently multimodal nature of music, which integrates both melody and lyrics, we also explore fundamental multimodal models in our experiments. Beyond providing basic binary classification outcomes, we delve deeper into model behaviour using multiple explainable Artificial Intelligence (XAI) tools, offering insights into their decision-making processes. Our analysis reveals that ResNet18 performs the best according to in-domain and out-of-domain tests. By providing a comprehensive comparison of systematic evaluation results and their interpretability, we propose several directions to inspire future research to develop more robust and effective detection methods for MGM. We provide our codes and some samples on Github repository.

Formal Explanations for Neuro-Symbolic AI

Sushmita Paul, Jinqiang Yu, Jip J. Dekker +2

Despite the practical success of Artificial Intelligence (AI), current neural AI algorithms face two significant issues. First, the decisions made by neural architectures are often prone to bias and brittleness. Second, when a chain of reasoning is required, neural systems often perform poorly. Neuro-symbolic artificial intelligence is a promising approach that tackles these (and other) weaknesses by combining the power of neural perception and symbolic reasoning. Meanwhile, the success of AI has made it critical to understand its behaviour, leading to the development of explainable artificial intelligence (XAI). While neuro-symbolic AI systems have important advantages over purely neural AI, we still need to explain their actions, which are obscured by the interactions of the neural and symbolic components. To address the issue, this paper proposes a formal approach to explaining the decisions of neuro-symbolic systems. The approach hinges on the use of formal abductive explanations and on solving the neuro-symbolic explainability problem hierarchically. Namely, it first computes a formal explanation for the symbolic component of the system, which serves to identify a subset of the individual parts of neural information that needs to be explained. This is followed by explaining only those individual neural inputs, independently of each other, which facilitates succinctness of hierarchical formal explanations and helps to increase the overall performance of the approach. Experimental results for a few complex reasoning tasks demonstrate practical efficiency of the proposed approach, in comparison to purely neural systems, from the perspective of explanation size, explanation time, training time, model sizes, and the quality of explanations reported.

Causality is vital for understanding true cause-and-effect relationships between variables within predictive models, rather than relying on mere correlations, making it highly relevant in the field of Explainable AI. In an automated decision-making scenario, causal inference methods can analyze the underlying data-generation process, enabling explanations of a model's decision by manipulating features and creating counterfactual examples. These counterfactuals explore hypothetical scenarios where a minimal number of factors are altered, providing end-users with valuable information on how to change their situation. However, interpreting a set of multiple counterfactuals can be challenging for end-users who are not used to analyzing raw data records. In our work, we propose a novel multi-step pipeline that uses counterfactuals to generate natural language explanations of actions that will lead to a change in outcome in classifiers of tabular data using LLMs. This pipeline is designed to guide the LLM through smaller tasks that mimic human reasoning when explaining a decision based on counterfactual cases. We conducted various experiments using a public dataset and proposed a method of closed-loop evaluation to assess the coherence of the final explanation with the counterfactuals, as well as the quality of the content. Results are promising, although further experiments with other datasets and human evaluations should be carried out.

Rater Cohesion and Quality from a Vicarious Perspective

Deepak Pandita, Tharindu Cyril Weerasooriya, Sujan Dutta +5

Human feedback is essential for building human-centered AI systems across domains where disagreement is prevalent, such as AI safety, content moderation, or sentiment analysis. Many disagreements, particularly in politically charged settings, arise because raters have opposing values or beliefs. Vicarious annotation is a method for breaking down disagreement by asking raters how they think others would annotate the data. In this paper, we explore the use of vicarious annotation with analytical methods for moderating rater disagreement. We employ rater cohesion metrics to study the potential influence of political affiliations and demographic backgrounds on raters' perceptions of offense. Additionally, we utilize CrowdTruth's rater quality metrics, which consider the demographics of the raters, to score the raters and their annotations. We study how the rater quality metrics influence the in-group and cross-group rater cohesion across the personal and vicarious levels.

The rapid expansion of Artificial Intelligence (AI) technologies has introduced both significant advancements and challenges, with diversity and inclusion (D&I) emerging as a critical concern. Addressing D&I in AI is essential to reduce biases and discrimination, enhance fairness, and prevent adverse societal impacts. Despite its importance, D&I considerations are often overlooked, resulting in incidents marked by built-in biases and ethical dilemmas. Analyzing AI incidents through a D&I lens is crucial for identifying causes of biases and developing strategies to mitigate them, ensuring fairer and more equitable AI technologies. However, systematic investigations of D&I-related AI incidents are scarce. This study addresses these challenges by identifying and understanding D&I issues within AI systems through a manual analysis of AI incident databases (AIID and AIAAIC). The research develops a decision tree to investigate D&I issues tied to AI incidents and populate a public repository of D&I-related AI incidents. The decision tree was validated through a card sorting exercise and focus group discussions. The research demonstrates that almost half of the analyzed AI incidents are related to D&I, with a notable predominance of racial, gender, and age discrimination. The decision tree and resulting public repository aim to foster further research and responsible AI practices, promoting the development of inclusive and equitable AI systems.

AuditMAI: Towards An Infrastructure for Continuous AI Auditing

Laura Waltersdorfer, Fajar J. Ekaputra, Tomasz Miksa +1

Artificial Intelligence (AI) Auditability is a core requirement for achieving responsible AI system design. However, it is not yet a prominent design feature in current applications. Existing AI auditing tools typically lack integration features and remain as isolated approaches. This results in manual, high-effort, and mostly one-off AI audits, necessitating alternative methods. Inspired by other domains such as finance, continuous AI auditing is a promising direction to conduct regular assessments of AI systems. The issue remains, however, since the methods for continuous AI auditing are not mature yet at the moment. To address this gap, we propose the Auditability Method for AI (AuditMAI), which is intended as a blueprint for an infrastructure towards continuous AI auditing. For this purpose, we first clarified the definition of AI auditability based on literature. Secondly, we derived requirements from two industrial use cases for continuous AI auditing tool support. Finally, we developed AuditMAI and discussed its elements as a blueprint for a continuous AI auditability infrastructure.

Explainability and transparency of AI systems are undeniably important, leading to several research studies and tools addressing them. Existing works fall short of accounting for the diverse stakeholders of the AI supply chain who may differ in their needs and consideration of the facets of explainability and transparency. In this paper, we argue for the need to revisit the inquiries of these vital constructs in the context of LLMs. To this end, we report on a qualitative study with 71 different stakeholders, where we explore the prevalent perceptions and needs around these concepts. This study not only confirms the importance of exploring the ``who'' in XAI and transparency for LLMs, but also reflects on best practices to do so while surfacing the often forgotten stakeholders and their information needs. Our insights suggest that researchers and practitioners should simultaneously clarify the ``who'' in considerations of explainability and transparency, the ``what'' in the information needs, and ``why'' they are needed to ensure responsible design and development across the LLM supply chain.